Hypothetical quesiton regarding "randomness"

Just to declare - I am not trying to push any particular theory here, it's just something I've thought about.

Take for example, random.org, it gives you a truly random password based on atmospheric noise - It's a useful website for truly random numbers.

Here's my example:

You take any measurement such as the recordings of water droplets hitting the sink from a dripping tap. You record it with ultra high precision cameras/censors and you do this for X amount of years, say 10,000.

From these recordings, you "crack" it like the brute force method a keygen for a trial program would work - the programs keep trying different algorithms (number sets) that fits the droplet recordings so that the program then finds the "key" and is able to then predict the next droplet position.

Some people may say 'well, that's easy - it would be all dependant on the drop speed and all the other variables, right?" Well, not really because the universe has obviously a lot of probabilities - subatomic particles shake randomly about and the dripping tap example would always be different if you were able to record it with such high detail.

Perhaps a super computer would be needed with memory the size of planets, perhaps even bigger - but to me, if you did find the key, perhaps then it would be easy to predict other areas.

I realise there are some systems like this existing such as weather prediction patterns but I think that's going over board right now and this is why I feel that if you used all those machines and focused them on a little example, you could do it. Perhaps a more practical example would be the natural shaking's of an atom.

By the very definition of true randomness you can never predict the outcome. Atmospheric noise isn't truly random I don't think, but since it follows chaos theory the amount of processing power needed to "crack" it will get expontentially larger as the size of the system increases. inb4 kaini/chengod/chaosmachine

  On 4/25/2010 at 10:02 PM, Rabid said:

By the very definition of true randomness you can never predict the outcome. Atmospheric noise isn't truly random I don't think, but since it follows chaos theory the amount of processing power needed to "crack" it will get expontentially larger as the size of the system increases. inb4 kaini/chengod/chaosmachine

So your saying it wouldn't ever be possible?

you can't crack it. all you can do is give probabilities.

edit: yes joshier it is impossible by definition.

good thread; i'm still not sure how a random number is generated in a computer, because surely by definition it's impossible? i kept berating my old computer/coding expert boss on this and i think he opted out to shut me up by saying something like computers getting the initial random value externally; like an arbitrary value of electronic interference or CPU heat or something?

  On 4/25/2010 at 10:14 PM, idrn said:

good thread; i'm still not sure how a random number is generated in a computer, because surely by definition it's impossible? i kept berating my old computer/coding expert boss on this and i think he opted out to shut me up by saying something like computers getting the initial random value externally; like an arbitrary value of electronic interference or CPU heat or something?

  Quote

Perhaps you have wondered how predictable machines like computers can generate randomness. In reality, most random numbers used in computer programs are pseudo-random, which means they are a generated in a predictable fashion using a mathematical formula. This is fine for many purposes, but it may not be random in the way you expect if you're used to dice rolls and lottery drawings.

I'm gonna sample me some true white noise

  On 4/25/2010 at 9:56 PM, joshier said:

Take for example, random.org, it gives you a truly random password based on atmospheric noise - It's a useful website for truly random numbers.

cheers, i skimmed over this bit as i've been drinking but it's also relevant!

but yeah, what you're talking about, it's just too many variables to contemplate and i don't get what you mean by 'key'? surely to predict it accurately, you'd have to isolate that tap from every conceivable variable except the one's you're aware of, then factor those one's in?

but theoretically you shouldn't be able to predict every fucking little thing, because on the bigger scale this comes down to the core idea of whether or not we have any control over anything/ it's not just cause and effect. if there isn't any randomness then i simply cannot be held responsible for molesting your hamster with a twix. and i don't what to be empowered by such ideas.

maybe even the question 'why are we here' is so random that it makes completely no sense...

I think key = algorithm?

as others have said, there is no way to predict truly random numbers, that's why they're truly random.

if you manage to find an algorithm that works, then you have a non-random source.

  On 4/25/2010 at 10:14 PM, idrn said:

good thread; i'm still not sure how a random number is generated in a computer, because surely by definition it's impossible? i kept berating my old computer/coding expert boss on this and i think he opted out to shut me up by saying something like computers getting the initial random value externally; like an arbitrary value of electronic interference or CPU heat or something?

Random number generators in computers are realy algorithms that take in an initial value and turn it into another number. So if you give it the same initial value (or "seed") twice it will give the same output twice. Usually they use something like computer time in milliseconds as the seed.

there exist a lot of different methods computers use to simulate random samples. they are quite good these days, and pretty much indistinguishable from true randomness.

  On 4/25/2010 at 10:14 PM, idrn said:

good thread; i'm still not sure how a random number is generated in a computer, because surely by definition it's impossible? i kept berating my old computer/coding expert boss on this and i think he opted out to shut me up by saying something like computers getting the initial random value externally; like an arbitrary value of electronic interference or CPU heat or something?

old programming languages used to generate it based on the number of seconds to midnight. not very random at all, but serviceable.

most "random number generators" are like this - they can get more complex and less predictable, but since they are based on natural inputs, they are reproducible given those inputs.

random number

I think essentially what youre talking about is that things which could be said to be random may actually be predictable if you had enough historical data or information about the system.

this page offers "truly" random numbers via radioactive decay:

https://www.fourmilab.ch/hotbits/secure_generate.html

stochastic modelling. something i was actually quite good at, for a change.

also, this might answer some thing joshier:

https://www.fourmilab.ch/documents/gtpp/

  On 4/26/2010 at 12:58 AM, Coalbucket PI said:

I think essentially what youre talking about is that things which could be said to be random may actually be predictable if you had enough historical data or information about the system.

and so statistics was born.

what if the bias lies in a random direction?

you can only talk about random numbers for so long before religion or philosophy come into the discussion . . ahh science.

have fun guys

  On 4/26/2010 at 3:59 AM, Mesh Gear Fox said:

I'm going on the logic that everything in nature requires some form of mathematics no matter how complex, otherwise it is illogical and hence inexistent.

probability is a field of mathematics too.

  On 4/26/2010 at 6:06 AM, Mesh Gear Fox said:

  On 4/26/2010 at 6:02 AM, GORDO said:

  On 4/26/2010 at 3:59 AM, Mesh Gear Fox said:

I'm going on the logic that everything in nature requires some form of mathematics no matter how complex, otherwise it is illogical and hence inexistent.

 

probability is a field of mathematics too.

true but probability is more of a prediction rather than something quantifiable.

not at all.

there's this thing called (quantitative) risk management, for instance.

are you trying to say that the world is deterministic? if so, consider quantum mechanics, in this realms things can't be but described in terms of probabilities until they are observed.

*brain explodes*